0

My newly built computer freezes randomly. When it freezes, it doesn't respond to keyboards or mouses, so REISUB or ctrl+alt+f1 (or f2,f3.. etc) are of no use.

I had tried

  • different graphic card drivers: the (fedora/ubuntu/win) defaults, the latest driver from Nvidia
  • different OS: Ubuntu 14.04 (with UEFI boot), Fedora 23 (tried with or w.o. UEFI boot), Windows 10
  • updating kernel
  • swap memory stick: tried use only one stick, and placing it in different slots
  • placing the graphic card in other PCI slots
  • updating the bios from ASRocks
  • disable firefox/chrome plugins/extensions. (at first I suspect the freeze is caused by browsers)

and the freeze still continues to occur.

I also did the following tests and they all report no errors:

  • MemTest86: It ran for 16 hours and pass all 4 tests. Results.
  • Glmark2 (to see how the graphic card performs. scores: 12688)
  • Intel Processor Diagnostic Tool in Windows 10
  • I also monitor the temperature and it remains below 60c

Configuration:

  • Intel® Core™ i7-5960X
  • ASRock X99X Killer LGA 2011-v3 Intel X99 SATA 6Gb/s USB 3.0 ATX Intel Motherboard
  • EVGA GeForce GTX 980
  • SAMSUNG 850 EVO 2.5" 500GB SSD
  • G.SKILL Ripjaws 4 Series 32GB (4 x 8GB) 288-Pin DDR4
  • CORSAIR CX series CX600M 600W

I know this configuration looks like a gaming computer but I've never played a game on it or overclocked. The funniest finding in this frustrated situation is that the computer freezes at the most unexpected moments: It can freeze when I'm just using the file explorer + terminal. It sometimes freezes after watching YouTube for 30 minutes. But it never freezes when running diagnose tools. It also didn't freeze when I were running some python homework code and all 16 cpu-cores reach 100% usage.

Question:

Since the freezing happens across OS, I'm convinced that some of my hardware is broken. Below are the errors that mcelog shows, repeatedly. Can someone tells which part of my hardware is broken from the log? Or how can I understand it better?

Thanks!

Jan 29 09:26:24 fedora23 mcelog[1004]: Hardware event. This is not a software error.
Jan 29 09:26:24 fedora23 mcelog[1004]: MCE 0
Jan 29 09:26:24 fedora23 mcelog[1004]: CPU 0 BANK 17
Jan 29 09:26:24 fedora23 mcelog[1004]: MISC 8cf00031e0000086 ADDR 5f000000
Jan 29 09:26:24 fedora23 mcelog[1004]: TIME 1454088380 Fri Jan 29 09:26:20 2016
Jan 29 09:26:24 fedora23 mcelog[1004]: MCG status:
Jan 29 09:26:24 fedora23 mcelog[1004]: MCi status:
Jan 29 09:26:24 fedora23 mcelog[1004]: Error overflow
Jan 29 09:26:24 fedora23 mcelog[1004]: Uncorrected error
Jan 29 09:26:24 fedora23 mcelog[1004]: MCi_MISC register valid
Jan 29 09:26:24 fedora23 mcelog[1004]: MCi_ADDR register valid
Jan 29 09:26:24 fedora23 mcelog[1004]: Processor context corrupt
Jan 29 09:26:24 fedora23 mcelog[1004]: MCA: Generic CACHE Level-2 Eviction Error
Jan 29 09:26:24 fedora23 mcelog[1004]: STATUS ee2000000004017a MCGSTATUS 0
Jan 29 09:26:24 fedora23 mcelog[1004]: MCGCAP c16 APICID 0 SOCKETID 0
Jan 29 09:26:24 fedora23 mcelog[1004]: CPUID Vendor Intel Family 6 Model 63
Jan 29 09:26:24 fedora23 mcelog[1004]: Hardware event. This is not a software error.
Jan 29 09:26:24 fedora23 mcelog[1004]: MCE 1
Jan 29 09:26:24 fedora23 mcelog[1004]: CPU 0 BANK 18
Jan 29 09:26:24 fedora23 mcelog[1004]: MISC 1cf00031e0000086 ADDR 5f100040
Jan 29 09:26:24 fedora23 mcelog[1004]: TIME 1454088380 Fri Jan 29 09:26:20 2016
Jan 29 09:26:24 fedora23 mcelog[1004]: MCG status:
Jan 29 09:26:24 fedora23 mcelog[1004]: MCi status:
Jan 29 09:26:24 fedora23 mcelog[1004]: Error overflow
Jan 29 09:26:24 fedora23 mcelog[1004]: Uncorrected error
Jan 29 09:26:24 fedora23 mcelog[1004]: MCi_MISC register valid
Jan 29 09:26:24 fedora23 mcelog[1004]: MCi_ADDR register valid
Jan 29 09:26:24 fedora23 mcelog[1004]: Processor context corrupt
Jan 29 09:26:24 fedora23 mcelog[1004]: MCA: Generic CACHE Level-2 Eviction Error
Jan 29 09:26:24 fedora23 mcelog[1004]: STATUS ee2000000004017a MCGSTATUS 0
Jan 29 09:26:24 fedora23 mcelog[1004]: MCGCAP c16 APICID 0 SOCKETID 0
Jan 29 09:26:24 fedora23 mcelog[1004]: CPUID Vendor Intel Family 6 Model 63
0

I had such problem long time ago. Check your memory (RAM) using memory test bundled with fedora and ubuntu DVD.

  • The memory test bundled with fedora and ubuntu liveCD is called memtest86+. It does not support UEFI boot. So alternatively I did memtest86, as suggested in this link, and no error is found. – Since1996 Jan 30 '16 at 7:35

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.